【发布时间】：2022-07-05 00:47:04
【问题描述】：
我是 pytorch 的初学者。我正在尝试进行基于方面的情绪分析。我正面临主题中提到的错误。我的代码如下：我请求帮助以解决此错误。提前致谢。我将分享整个代码和错误堆栈。
!pip install transformers
import transformers
from transformers import BertModel, BertTokenizer, AdamW, get_linear_schedule_with_warmup
import torch
import numpy as np
import pandas as pd
import seaborn as sns
from pylab import rcParams
import matplotlib.pyplot as plt
from matplotlib import rc
from sklearn.model_selection import train_test_split
from sklearn.metrics import confusion_matrix, classification_report
from collections import defaultdict
from textwrap import wrap
from torch import nn, optim
from torch.utils.data import Dataset, DataLoader
%matplotlib inline
%config InlineBackend.figure_format='retina'
sns.set(style='whitegrid', palette='muted', font_scale=1.2)
HAPPY_COLORS_PALETTE = ["#01BEFE", "#FFDD00", "#FF7D00", "#FF006D", "#ADFF02", "#8F00FF"]
sns.set_palette(sns.color_palette(HAPPY_COLORS_PALETTE))
rcParams['figure.figsize'] = 12, 8
RANDOM_SEED = 42
np.random.seed(RANDOM_SEED)
torch.manual_seed(RANDOM_SEED)
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
df = pd.read_csv("/Users/user1/Downloads/auto_bio_copy.csv")
我正在导入一个 csv 文件，其内容和标签如下所示：
df.head()
content label
0 I told him I would leave the car and come back... O O O O O O O O O O O O O O O O O O O O O O O ...
1 I had the ignition interlock device installed ... O O O B-Negative I-Negative I-Negative O O O O...
2 Aug. 23 or 24 I went to Walmart auto service d... O O O O O O O B-Negative I-Negative I-Negative...
3 Side note This is the same reaction I 'd gotte... O O O O O O O O O O O O O O O O O O O O O O O ...
4 Locked out of my car . Called for help 215pm w... O O O O O O O O O O O O O O O O O B-Negative O...
df.shape
(1999, 2)
我将标签值转换为整数，如下所示： O=0(0), B-Positive=1, I-Positive=2, B-Negative=3, I-Negative=4, B-Neutral=5, I-Neutral=6, B-Mixed=7, I -混合=8
df['label'] = df.label.str.replace('O', '0')
df['label'] = df.label.str.replace('B-Positive', '1')
df['label'] = df.label.str.replace('I-Positive', '2')
df['label'] = df.label.str.replace('B-Negative', '3')
df['label'] = df.label.str.replace('I-Negative', '4')
df['label'] = df.label.str.replace('B-Neutral', '5')
df['label'] = df.label.str.replace('I-Neutral', '6')
df['label'] = df.label.str.replace('B-Mixed', '7')
df['label'] = df.label.str.replace('I-Mixed', '8')
接下来，将字符串转换为整数列表如下：
df['label'] = df['label'].str.split(' ').apply(lambda s: list(map(int, s)))
df.head()
content label
0 I told him I would leave the car and come back... [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...
1 I had the ignition interlock device installed ... [0, 0, 0, 3, 4, 4, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...
2 Aug. 23 or 24 I went to Walmart auto service d... [0, 0, 0, 0, 0, 0, 0, 3, 4, 4, 4, 0, 0, 0, 0, ...
3 Side note This is the same reaction I 'd gotte... [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...
4 Locked out of my car . Called for help 215pm w... [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...
PRE_TRAINED_MODEL_NAME = 'bert-base-cased'
tokenizer = BertTokenizer.from_pretrained(PRE_TRAINED_MODEL_NAME)
token_lens = []
for txt in df.content:
tokens = tokenizer.encode_plus(txt, max_length=512, add_special_tokens=True, truncation=True, return_attention_mask=True)
token_lens.append(len(tokens))
MAX_LEN = 512
class Auto_Bio_Dataset(Dataset):
def __init__(self, contents, labels, tokenizer, max_len):
self.contents = contents
self.labels = labels
self.tokenizer = tokenizer
self.max_len = max_len
def __len__(self):
return len(self.contents)
def __getitem__(self, item):
content = str(self.contents[item])
label = self.labels[item]
encoding = self.tokenizer.encode_plus(
content,
add_special_tokens=True,
max_length=self.max_len,
return_token_type_ids=False,
#padding='max_length',
pad_to_max_length=True,
truncation=True,
return_attention_mask=True,
return_tensors='pt'
)
return {
'content_text': content,
'input_ids': encoding['input_ids'].flatten(),
'attention_mask': encoding['attention_mask'].flatten(),
'labels': torch.tensor(label)
}
df_train, df_test = train_test_split(
df,
test_size=0.1,
random_state=RANDOM_SEED
)
df_val, df_test = train_test_split(
df_test,
test_size=0.5,
random_state=RANDOM_SEED
)
df_train.shape, df_val.shape, df_test.shape
((1799, 2), (100, 2), (100, 2))
def create_data_loader(df, tokenizer, max_len, batch_size):
ds = Auto_Bio_Dataset(
contents=df.content.to_numpy(),
labels=df.label.to_numpy(),
tokenizer=tokenizer,
max_len=max_len
)
return DataLoader(
ds,
batch_size=batch_size,
num_workers=2
)
BATCH_SIZE = 16
train_data_loader = create_data_loader(df_train, tokenizer, MAX_LEN, BATCH_SIZE)
val_data_loader = create_data_loader(df_val, tokenizer, MAX_LEN, BATCH_SIZE)
test_data_loader = create_data_loader(df_test, tokenizer, MAX_LEN, BATCH_SIZE)
data = next(iter(train_data_loader))
data.keys()
错误如下：
---------------------------------------------------------------------------
RuntimeError Traceback (most recent call last)
<ipython-input-71-e0a71018e473> in <module>
----> 1 data = next(iter(train_data_loader))
2 data.keys()
~/opt/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py in __next__(self)
528 if self._sampler_iter is None:
529 self._reset()
--> 530 data = self._next_data()
531 self._num_yielded += 1
532 if self._dataset_kind == _DatasetKind.Iterable and \
~/opt/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py in _next_data(self)
1222 else:
1223 del self._task_info[idx]
-> 1224 return self._process_data(data)
1225
1226 def _try_put_index(self):
~/opt/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py in _process_data(self, data)
1248 self._try_put_index()
1249 if isinstance(data, ExceptionWrapper):
-> 1250 data.reraise()
1251 return data
1252
~/opt/anaconda3/lib/python3.7/site-packages/torch/_utils.py in reraise(self)
455 # instantiate since we don't know how to
456 raise RuntimeError(msg) from None
--> 457 raise exception
458
459
RuntimeError: Caught RuntimeError in DataLoader worker process 0.
Original Traceback (most recent call last):
File "/Users/namrathabhandarkar/opt/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/worker.py", line 287, in _worker_loop
data = fetcher.fetch(index)
File "/Users/namrathabhandarkar/opt/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 52, in fetch
return self.collate_fn(data)
File "/Users/namrathabhandarkar/opt/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/collate.py", line 157, in default_collate
return elem_type({key: default_collate([d[key] for d in batch]) for key in elem})
File "/Users/namrathabhandarkar/opt/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/collate.py", line 157, in <dictcomp>
return elem_type({key: default_collate([d[key] for d in batch]) for key in elem})
File "/Users/namrathabhandarkar/opt/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/collate.py", line 138, in default_collate
return torch.stack(batch, 0, out=out)
RuntimeError: stack expects each tensor to be equal size, but got [157] at entry 0 and [154] at entry 1
我在一些github帖子中发现这个错误可能是因为批量大小，所以我将批量大小更改为8，然后错误如下：
BATCH_SIZE = 8
train_data_loader = create_data_loader(df_train, tokenizer, MAX_LEN, BATCH_SIZE)
val_data_loader = create_data_loader(df_val, tokenizer, MAX_LEN, BATCH_SIZE)
test_data_loader = create_data_loader(df_test, tokenizer, MAX_LEN, BATCH_SIZE)
data = next(iter(train_data_loader))
data.keys()
RuntimeError Traceback (most recent call last)
<ipython-input-73-e0a71018e473> in <module>
----> 1 data = next(iter(train_data_loader))
2 data.keys()
~/opt/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py in __next__(self)
528 if self._sampler_iter is None:
529 self._reset()
--> 530 data = self._next_data()
531 self._num_yielded += 1
532 if self._dataset_kind == _DatasetKind.Iterable and \
~/opt/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py in _next_data(self)
1222 else:
1223 del self._task_info[idx]
-> 1224 return self._process_data(data)
1225
1226 def _try_put_index(self):
~/opt/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py in _process_data(self, data)
1248 self._try_put_index()
1249 if isinstance(data, ExceptionWrapper):
-> 1250 data.reraise()
1251 return data
1252
~/opt/anaconda3/lib/python3.7/site-packages/torch/_utils.py in reraise(self)
455 # instantiate since we don't know how to
456 raise RuntimeError(msg) from None
--> 457 raise exception
458
459
RuntimeError: Caught RuntimeError in DataLoader worker process 0.
Original Traceback (most recent call last):
File "/Users/namrathabhandarkar/opt/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/worker.py", line 287, in _worker_loop
data = fetcher.fetch(index)
File "/Users/namrathabhandarkar/opt/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 52, in fetch
return self.collate_fn(data)
File "/Users/namrathabhandarkar/opt/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/collate.py", line 157, in default_collate
return elem_type({key: default_collate([d[key] for d in batch]) for key in elem})
File "/Users/namrathabhandarkar/opt/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/collate.py", line 157, in <dictcomp>
return elem_type({key: default_collate([d[key] for d in batch]) for key in elem})
File "/Users/namrathabhandarkar/opt/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/collate.py", line 137, in default_collate
out = elem.new(storage).resize_(len(batch), *list(elem.size()))
RuntimeError: Trying to resize storage that is not resizable
我不确定是什么导致了第一个错误（主题中提到的那个）。我在我的代码中使用填充和截断，但是错误。
非常感谢任何解决此问题的帮助。
提前致谢。
【问题讨论】：
标签： pytorch sentiment-analysis pytorch-dataloader